P12-35/2.8 & P12-60/3.5 vignetting

threeOh

Mu-43 Regular
Joined
Aug 22, 2019
Messages
53
Both of these lenses vignette noticeably, at pretty much all apertures. By 14mm its sorta OK. I feel like I paid for 12's and got 14's. Of the very few lens "reviews" that even mentioned vignetting, only one, out of at least a dozen, seemed disturbed about it. I use the wide end a lot. And when I do its usually in the high desert areas out west where I'm taking a lot of shots at a lot of angles of a lot of scenery. No lens changing in that environment. Prefer not to carry a second body and purchase a 12mm prime. Even after my normal aggressive culling, dealing with that in post has turned out to be a real bummer. As such, I'm thinking of ditching both lenses and getting the 14-140 for my next trip and just use the 12-32 where I'd normally use the 12-35. I'm basically a prime shooter so losing some zooms is not a big deal -- except when I'm traveling -- at least 4 months a year.

This seems an internal hardware issue. Vignettes even without filters or hoods. Included lens corrections do nothing at all about it. Checked Lightroom (latest version, Classic), Lens Corrections is turned on. Turn it off, no change. Tried Lightroom Vignetting tool but, as usual, varying backgrounds often make that a tedious job with poor results. Could crop top and bottom but not sides.

So, how are you folks dealing with this vignetting?

If I'm off base, let me know. I must admit I don't see any critical comments about this.
 

wjiang

Mu-43 Legend
Joined
Sep 7, 2013
Messages
7,518
Location
Christchurch, New Zealand
I find most of my lenses vignette to some extent, though I've noticed it the most with Panasonic telephotos. Yeah it can be a pain occasionally, but most of the time I end up adding a post-crop vignette for effect anyway so it's usually fine.
 

PakkyT

Mu-43 Hall of Famer
Joined
Jun 20, 2015
Messages
2,893
Location
New England
I recently sold my Panny 12-35/2.8 but in the years I noticed it, I never noticed vignetting with mine. Also I used it with the Oly E-M1.1 camera, but I can not imagine this would be a condition only on Panny bodies. To be clear I assume you mean a slight darkening of the corners (rather than see the edge of the lens physically in the field of view, right?).

It could also be that I just don't notice it like others might? Or as @wjiang mentioned, since I too tend to do a bit of straightening in post (and cropping when needed) perhaps I am editing it out without noticing?
 

bargainguy

Mu-43 Regular
Joined
Apr 21, 2015
Messages
34
Location
Milwaukee, WI
Real Name
Don
I see you have Lightroom. Do you have Photoshop?

In my PS version (20.0.6), from the drop-down menu: Filter > Lens Correction.

This takes you to a separate dialog box, where there are two tabs, Auto Correction and Custom.

Under the Auto Correction tab, you can check the Vignette box to auto-correct.

Feel like doing it manually? Under the Custom tab, you have a Vignette slider control with a range of -100 to +100, with 0 being neutral.
 

Peadingle

Mu-43 Regular
Joined
Dec 3, 2018
Messages
95
Just got the 12-60 3.5/5.6, and have noticed vignetting when processing in ACR but not Photolabs2
 

threeOh

Mu-43 Regular
Joined
Aug 22, 2019
Messages
53
I appreciate the feedback. I’ve been, and will continue to be, out of town for the next few days so, no access to my images.

Responding to some feedback in a collective manner:
I can’t see myself adding a vignette. Just don’t see me expanding what few images have it applied now. Will do a few affected images and maybe I’ll be surprised.

I'm going to say it’s quite bad. Like over a stop to 2 stops. I’m going to shoot a better controlled set of images but, right now I’d suggest black rather than grey. But I’ll admit my quick look this morning showed inconsistencies.

I believe Lightroom has the same vignetting module as PS. If not I can install my paid for every month never downloaded copy. Been a while, hopefully it’s a bit more efficient with resources these days.

I'm new to m4/3 and am still dipping my toes in some areas. Last week I saw a DXO Black Friday discount that was paramount to giving the product away. Went to get it today but too late. I was thinking DXO has a pretty good handle on lens calibrations and Photolab appears to be highly thought of for m4/3. Thanks for pushing me to download a trial and running a few images through it.

At this point my current potential solution is to compose with more sky than I want so I can crop the top, fix the bottoms and leave the sides.

I suspect some of my reaction is due to 40 years ago I bought a Nikon zoom. It lasted one shoot. Too big, too slow, too meh. Other than some kit Fuji's that came along for the ride, I've been shooting primes. Zooms are nice for travel so I now find myself with 4 of them. Used only the 12-60 so far.
 

ColinM

Mu-43 Rookie
Joined
Apr 27, 2012
Messages
20
Location
Llandudno, UK
Real Name
Colin
It sounds as if you have a bad copy of that 12-60mm, because I've never seen anything like it. I just put my copy on my E-M1 II and took a couple of quick shots. Using the Color Sampler Tool in ACR there seems to be about 1/4 stop falloff at the sides and maybe 1/2 stop in the extreme corners at 12mm. I would think that is acceptable for a wide zoom like this. Like wjiang I usually vignette slightly in post so it just isn't an issue for me.
 

threeOh

Mu-43 Regular
Joined
Aug 22, 2019
Messages
53
Lenstip tests vignetting at 12mm, f3.5 at -1.78 stops. That’s consistent with what I’m seeing. I doubt this is a bad copy issue. My guess at the moment is it’s a basic design issue where Panasonic decided to go tight with the size of the barrel.

I do find it interesting though that perhaps 80+% of the reviewers/testers don’t even mention vignetting. Nor do some forum contributors. My corners are reasonably even so I don’t see it being caused by a wildly decentered lens. Though I will shoot a chart once home.
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,397
I'm going to say it’s quite bad. Like over a stop to 2 stops.
Vignetting is routinely measured in lens tests and 1-2 stops wide open is entirely normal for wides in any system. As such, I'd suggest perhaps more of a focus on measurement than subjective reviews would be helpful. Or, if it's important to continue thinking of ~1.5 stops of vignetting as quite bad, then avoiding measurements of ultrawides, wide primes, and certain lenses such as the EF-M 15-45 would probably be a good choice. ;)

Depending on it's exact intent, the claim of "pretty much all apertures" made in this thread's first post may be poorly supported by measurements of the 12-35 and 12-60. So perhaps some methodological review would be appropriate. I don't use Adobe tools but they should be honouring the embedded correction in the .cr2 if that's what you're working from. If not, transcoding to DNG and checking the FixVignetteRadial opcode should provide a toolchain check.

This seems an internal hardware issue.
One possibility is the core concept you're looking for is cos⁴() falloff, which is routinely described in introductions to lens properties. So lots of options to choose from if one's so inclined. It's a simplification but is broadly applicable and often referred to as a law (Gardner 1947).

For example, some similar lenses measured to have one to two stops of vignetting are
  • Canon 15-85
  • Fuji 16-50
  • Fuji 16-55
  • Nikon 16-85
  • Olympus 12-40
  • Olympus 12-50
  • Olympus 12-100
  • Panasonic-Leica 12-60
  • Sony 16-50 f/2.8
  • Zeiss 16-70
However, it's a bit unclear from the above whether this is the correct range of stops.

So, how are you folks dealing with this vignetting?
Not having to create lensfun profiles is one of the reasons I like SOOCs, though I'm not aware of any manufacturer which fully corrects vignetting (0.5-1 stop after correction when wide open is normal for wides). I actually don't think I've ever needed to fall back to raw with the 12-60. On the rare occasions I've wanted vignette correction on raws, which is usually with the 100-300 II, I've used both darktable and RawTherapee.

Given IBIS and OIS vignette reduction isn't necessarily trivially automatable as the position of the correction groups and sensor shifts the location of the optical axis. There are also quite a few threads here about inconsistent mechanical vignetting due to IBIS shifting corners of the sensor close to the edge of or out of the image circle. Anecdotally that seems to happen more with Olympus than Panasonic (not sure about Fuji, Nikon, or Sony's implementations) and could be argued to be more of a feature than a bug.
 
Last edited:

sgt08

Mu-43 All-Pro
Joined
Mar 12, 2011
Messages
1,201
Location
Hong Kong
Real Name
Scott
I've also been using the 12-60 for travel/scenic photos and have not noticed anywhere near that much vignetting at 12mm, although I'm more likely to shoot at f/5.6-8 range for those kinds of shots. I wonder if the Lenstip results have something to do with their testing process, like perhaps their JPGs are coming from the same process that they use to measure RAW distortion without the corrections that are written into the RAW file (as @archaeopteryx mentioned)? In LR even if you turn off Lens Corrections it still applies "built in lens profile" corrections from the RAW. But it sounds like you are using LR, so if this vignetting was supposed to be corrected and you're still seeing it in your photos…I don't know what to think!

If it helps, I just did a quick skim through my LR catalog for images at 12mm in the f/3.5-4 range, unfortunately I don't have anything really ideal for checking vignetting but here are a couple that have some corner and side regions where I think you would notice that much falloff. I did a straight JPG export from uncropped RAW with everything disabled/zeroed out, these were taken with a GX85. Will be interesting to see how your controlled test shots compare.
 

Attachments

bargainguy

Mu-43 Regular
Joined
Apr 21, 2015
Messages
34
Location
Milwaukee, WI
Real Name
Don
I'm thinking the 12-60 is not an ideal lens for you if:

1. You like the quality of primes, and

2. The vignetting is severe enough that you don't even want to deal with it in post.

Having to include more sky just to be able to crop back and avoid vignetting - that's a huge kludge, and gets old real fast.

Why not a small WA prime? The Oly 12/2 comes to mind. I don't own one, but as far as I remember, doesn't seem to have serious vignette issues from the samples I've seen.

I'll throw in a plug for the Panny 7-14. Got this lens a few months ago and was expecting a wallop of vignetting, esp. at the 7mm end, but to my surprise, almost none!

Don't know if that's due to a built-in lens correction profile or just the optical design, but either way, it works for me.
 
Last edited:

threeOh

Mu-43 Regular
Joined
Aug 22, 2019
Messages
53
I read this last night and spent the morning on it.

Vignetting is routinely measured in lens tests and 1-2 stops wide open is entirely normal for wides in any system. As such, I'd suggest perhaps more of a focus on measurement than subjective reviews would be helpful. Or, if it's important to continue thinking of ~1.5 stops of vignetting as quite bad, then avoiding measurements of ultrawides, wide primes, and certain lenses such as the EF-M 15-45 would probably be a good choice. ;)

Coming from fast (f1.4 primes) I did not expect to see vignetting at f3.5. I was wrong. I continue to feel the lens is fine. I assume your "avoiding measurements" should read simply "avoiding". To which I'll respond its too early for that. I bought this system for travel and will simply have to get used to this lens. Looking at my trip to Canyonlands, 12mm turned out to be my most used focal length. Wider and the perspective distortion becomes too difficult for me.

Depending on it's exact intent, the claim of "pretty much all apertures" made in this thread's first post may be poorly supported by measurements of the 12-35 and 12-60.

The only measurements I found of the 12-60, along with images, was the Lens Tip tests. After adjustments (SOOC jpeg). At f8 vignetting is -.8ev. Thats readily visible in the corners with a light sky, be it blue or grey. It also creates a large, though subdued, vignetted area that's creeps well into the image. I don't go past f8 with m4/3. Don't know what its like beyond that. Call me picky, call me spoiled but I see it and I'd prefer not to. And no I am not selling prints. This is purely amateur travel photography hoping I come back with more than a slide show. For those, I'd prefer not having to correct uneven lighting.

T
ook many shots of the sky and a blank wall this morning with all sorts of filters and hoods I generally use when filters are preferred (not often but were used in the shots out west that led to my initial post). Agree, without the ancient (thick) 58mm PL filter I was using, the vignetting is almost gone mostly acceptable by f8. But at f3.5 and 5.6 I need/want to deal with it. Unfortunately there's insufficient room to mount a 62mm step-up + 62mm pl with the stock hood is mounted. My usual screw-on hood exacerbates the vignetting. Will need to look at alternate hoods or another PL.

One possibility is the core concept you're looking for is cos⁴() falloff, which is routinely described in introductions to lens properties. So lots of options to choose from if one's so inclined. It's a simplification but is broadly applicable and often referred to as a law (Gardner 1947).

This led to some interesting reactions. After the second read I embraced what Gardner said, "To determine the size of either entrance or exit pupil by computation is a tedious process." Conceptually I think I'm there. I don't need to understand the math or purchase an equipped lab and conduct experiments. But thank you as the paper led to other reads that were also new information for me.

For example, some similar lenses measured to have one to two stops of vignetting are
  • Canon 15-85
  • Fuji 16-50
  • Fuji 16-55
  • Nikon 16-85
  • Olympus 12-40
  • Olympus 12-50
  • Olympus 12-100
  • Panasonic-Leica 12-60
  • Sony 16-50 f/2.8
  • Zeiss 16-70
However, it's a bit unclear from the above whether this is the correct range of stops.

I owned the Fuji 16-50 and felt it was a 17 at best. Distortion correction cropped out the vignetting. A degree of correction which appears unnecessary with the P12-60.

Not having to create lensfun profiles is one of the reasons I like SOOCs, though I'm not aware of any manufacturer which fully corrects vignetting (0.5-1 stop after correction when wide open is normal for wides). I actually don't think I've ever needed to fall back to raw with the 12-60. On the rare occasions I've wanted vignette correction on raws, which is usually with the 100-300 II, I've used both darktable and RawTherapee.

I've been shooting jpegs for quite some time with Fuji. Adjusting the GX85 jpegs to my preference has yet to be done. I know I'm not there with the defaults and adjustments will take time. From a lens corrections pov, again with the P12-60, the jpeg offers essentially no advantage vs the raw file. The Panasonic corrections provided to Adobe result in essentially the same corners. Makes sense. Not fully correcting when wide open I assume is that the corrections are fixed across all fl's. If that's the case, makes sense and is reflected with this lens.

Given IBIS and OIS vignette reduction isn't necessarily trivially automatable as the position of the correction groups and sensor shifts the location of the optical axis. There are also quite a few threads here about inconsistent mechanical vignetting due to IBIS shifting corners of the sensor close to the edge of or out of the image circle. Anecdotally that seems to happen more with Olympus than Panasonic (not sure about Fuji, Nikon, or Sony's implementations) and could be argued to be more of a feature than a bug.

Understood, would expect its more likely in low light and dark corners. Sort of an environmental vignetting correction. There are inconsistencies in my Canyonlands images. But I'm not going into IBIS and OIS effects. Life's too short.
Again, I appreciate the feedback. If anything it caused me to revisit how I'm using the lens, if I want the lens and some changes I need to make.
 

threeOh

Mu-43 Regular
Joined
Aug 22, 2019
Messages
53
I'm thinking the 12-60 is not an ideal lens for you if:

1. You like the quality of primes, and

2. The vignetting is severe enough that you don't even want to deal with it in post.

Having to include more sky just to be able to crop back and avoid vignetting - that's a huge kludge, and gets old real fast.

Why not a small WA prime? The Oly 12/2 comes to mind. I don't own one, but as far as I remember, doesn't seem to have serious vignette issues from the samples I've seen.

I'll throw in a plug for the Panny 7-14. Got this lens a few months ago and was expecting a wallop of vignetting, esp. at the 7mm end, but to my surprise, almost none!

Don't know if that's due to a built-in lens correction profile or just the optical design, but either way, it works for me.
Agree, it's not. Most of my photography revolves around shooting fixed focal camera's or lenses. However, I bought m4/3 for travel and effective stabilization (in my 70's and Panasonics dual stabilization has been a blessing). So maybe its time I learned to use a zoom. This is not my first tome with zooms and they have seldom lasted long. For 25 years I kept a place in Europe and could spend my entire summer shooting a 24 or 50. Now permanently in the USA, I'm discovering the National Parks are not too kind to my single focal length approach. So, more patience this time.

In spite of the fact a 24 has been one of my principal fl's for 40+ years, its too soon to pick up a wide prime. The 7-14 is not a solution. I tend to shoot wide and close, single object, usually with depth, is the main focus. Perspective distortion below 12mm fl is difficult for me to handle well.

For now I deal with what I have. All this kit is brand new this year. Need to learn how to deal with what I have at this point.
 

BPCS

Mu-43 Regular
Joined
Oct 15, 2017
Messages
85
12mm and wider lenses produce a polarizing effect of the sky due to the extreme difference in angle of light from different parts of the sky. With a stop or two of lens vignetting produced if the lens is used wide open, this can result in several stops darker tops and corners of a sky. Lens profiles pretty much eliminate the lens optical vignetting, but the natural sky vignetting with ultra wide will remain. If you also use a polarisers with an ultra wide angle lens, you can get even more extreme shading and at an angle unrealistically across the frame, like a banding.

You will find that good lens profiles like from DXO PL 3 do work very well... get square to and close up to a large side of a building wall that is evenly illuminated and shoot it wide open. After appropriate processing (RAW in DXO PL3) and you will see 99% of lens vignetting is corrected (darkening towards corners gone). Lens profiles cannot correct nature... the darkening/ fading effect of light coming from greatly different angles from the sky. Our brain compensates and our eye doesn't see all this area at the same time, but a camera sensor does.
 

junkyardsparkle

haunted scrap heap
Joined
Nov 17, 2016
Messages
1,937
Location
Like, The Valley
For example, some similar lenses measured to have one to two stops of vignetting are
I can personally attest the same for the Oly f/1.8 primes up to the 45mm. One thing to keep in mind is that software correction is part of the m4/3 design philosophy generally, so this shouldn't really shock anybody... and, as mentioned already, in combination with IBIS this can sometimes produce a really dark corner or two. For evaluation/profiling purposes, it's important to disable IBIS.
Why not a small WA prime? The Oly 12/2 comes to mind. I don't own one, but as far as I remember, doesn't seem to have serious vignette issues from the samples I've seen.
I do own one, and it does. :D You may have seen OOC samples that were already corrected. This is one reason this discussion can get confusing; it's hard to know what people are really evaluating unless they state the case very clearly. AFAIK, even software like LR tries very hard not to display RAW files uncorrected when it has correction data for the lens in question, so even if somebody says they were "looking at RAW files, not JPEGs", well... take it with a grain of salt. Depending on the software you use, you might be seeing something different.

One more thing worth mentioning, just to further muddy the waters: the amount of vignetting can change quite a bit depending on focus distance... and unfortunately, it tends to be strongest at infinity, which is probably the opposite of what most people would want. It's another thing to keep in mind when trying to do meaningful comparisons. From what I've looked at (mostly small primes), internal-focus lenses seem to exhibit this to a greater degree than manual unit-focusing lenses, but there may very well be exceptions depending on the specific design.
 

threeOh

Mu-43 Regular
Joined
Aug 22, 2019
Messages
53
12mm and wider lenses produce a polarizing effect of the sky due to the extreme difference in angle of light from different parts of the sky.

That’s quite interesting. Somewhere above I either mentioned, or decided not to mention, some inconsistencies I observed with the Canyonlands images. Same day, same place, same sky, vignetting in some images, not in others. Assumed it was the cpl. I’ll try to revisit this early tomorrow morning shooting east and west. I was about to say I’m not shooting wide open, then I realized f3.5 is wide open with the P12-60.

You will find that good lens profiles like from DXO PL 3 do work very well... get square to and close up to a large side of a building wall that is evenly illuminated and shoot it wide open. After appropriate processing (RAW in DXO PL3) and you will see 99% of lens vignetting is corrected (darkening towards corners gone). Lens profiles cannot correct nature... the darkening/ fading effect of light coming from greatly different angles from the sky. Our brain compensates and our eye doesn't see all this area at the same time, but a camera sensor does.

Plan to trial DXO. Initially from a noise reduction pov and now lens profiles as well.

Thanks for more education.
 

archaeopteryx

Gambian sidling bush
Joined
Feb 25, 2017
Messages
1,397
Don't know if that's due to a built-in lens correction profile or just the optical design.
It's vignetting correction. Your definition of almost none might also overlap with @threeOh's definition of unacceptable, at least at 7mm and f/4. ;)

There is, perhaps, an interesting lens selection consideration underlying this thread. Choosing f/4-5.6ish zooms to reduce size, weight, and cost compared to faster zooms or groupings of primes removes the ability reduce vignetting by stopping down to f/4-5.6 from some faster aperture. Since the lenses are also sharpest at f/4-5.6 (for 16MP testing) this nominally introduces a diffraction. For traditional near-far wide angle compositions which aren't focus bracketed it isn't a limiting factor, though.

I can personally attest the same for the Oly f/1.8 primes up to the 45mm.
Yep. Same for rather a lot of primes if one's looking at falloff near their maximum apertures. I thought about getting into that above but, as this thread seems prompted by surprise over commonplace lens properties, it's my working assumption @threeOh isn't using primes this way. That seems consistent with primarily wide and near-far compositions.

internal-focus lenses seem to exhibit this to a greater degree than manual unit-focusing lenses
Moving to closer focus expands the image circle by extension in unit focusing, endocentric lenses. So there's both some motion of the dimmer parts of the image off the corners of the sensor and a reduction in effective aperture which stops the lens down slightly.

The internal focus designs I've looked at focus closer by moving a positive group forward, a negative group rearward, or combinations thereof for lenses with multiple focus groups (US20190265450A1 and US20190339497A1, for example). These are all anti-zoom motions, consistent with focus breathing and 9-18 measurement, but I'd need to simulate what's going on with falloff. There's unfortunately software license constraints there.

some changes I need to make
Those would be?

12mm and wider lenses produce a polarizing effect of the sky
[citation needed]
 

Growltiger

Mu-43 All-Pro
Joined
Mar 26, 2014
Messages
1,963
Location
UK
Another idea. Please try an experiment. Look at a JPG straight from the camera instead of the raw file you are processing currently. Does it show the same vignetting or is there no longer any vignetting?
The reason for the experiment is because the M43 standard requires the data from the chip in the lens to include both linearity corrections (e.g. barrel distortion) and vignetting correction. The camera applies those corrections automatically when it makes a JPG.
Lightroom should be applying the same corrections automatically - it gets that special M43 correction data from the raw file.
 

threeOh

Mu-43 Regular
Joined
Aug 22, 2019
Messages
53
Another idea. Please try an experiment. Look at a JPG straight from the camera instead of the raw file you are processing currently. Does it show the same vignetting or is there no longer any vignetting?
The reason for the experiment is because the M43 standard requires the data from the chip in the lens to include both linearity corrections (e.g. barrel distortion) and vignetting correction. The camera applies those corrections automatically when it makes a JPG.
Lightroom should be applying the same corrections automatically - it gets that special M43 correction data from the raw file.
Had done that. The correction is the same. Any difference in the strength of the vignetting (area being the same) is due to processing baked into the jpegs.

Reversing the correction for the raws shows a considerable amount of correction.

Downloaded a trial of DXO. Impressive processor, may spend more time with it. For the 12-60 their vignetting corrections are essentially the same as Adobe's. As are the crops to the sides. DXO does not appear to have a correction module for every lens. Perhaps the 12-60 is not covered and they're using Panasonic's. I forget if I saw a message or not. (DXO is on a clone. Going to bed will take priority over curiosity.)
 
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Mu-43 is a fan site and not associated with Olympus, Panasonic, or other manufacturers mentioned on this site.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Copyright © 2009-2019 Amin Forums, LLC
Top Bottom